Addressing LLM Diversity by Infusing Random Concepts

2026-01-26

AI Summary

This work addresses the limited output diversity commonly observed in large language models (LLMs) during generative tasks. To mitigate this issue, the authors propose a prompt engineering approach that injects task-irrelevant random words or sentences into the input prompt to stimulate more diverse model responses. A systematic evaluation protocol encompassing multiple diversity metrics is developed and applied across several mainstream LLMs. Experimental results demonstrate that the proposed method significantly enhances output diversity, particularly in enumeration-style tasks, thereby confirming its effectiveness and generalizability. This approach offers a simple yet efficient strategy for improving the diversity of LLM-generated content without requiring model retraining or architectural modifications.

Technology Category

• Natural Language Processing: Prompt Engineering / Prompting
• Machine Learning: Large Multimodal Models (LMMs)
• Planning, Routing, and Scheduling: Planning with Language Models

Application Category

• User Modeling, Personalization and Recommendation: Large Language Models (LLM) for user modeling and recommendation
• Economics, Online Markets and Human Computation: LLM based quality controls for crowd work
• Search and Retrieval-Augmented AI: Large language models for search

Abstract

Large language models (LLMs) are known to produce outputs with limited diversity. In this work, we study whether infusing random concepts in the prompts can improve the diversity of the generated outputs. To benchmark the approach, we design a systematic evaluation protocol which involves prompting an LLM with questions of the form"Name 10 Hollywood actors", and analyzing diversity measures of the resulting LLM outputs. Our experiments on multiple LLMs show that prepending random words/sentences unrelated to the prompt result in greater diversity in the outputs of LLMs. We believe that this promising result and the evaluation protocol opens up interesting avenues for future work, such as how infusing randomness into LLMs could be applied to other domains. Further, the evaluation protocol could also inspire research into benchmarking LLM diversity more systematically.

Problem

• LLM diversity
• output diversity
• large language models
• randomness infusion

Innovation

• LLM diversity
• random concept infusion
• prompt engineering
• diversity evaluation
• large language models